How Glance turns hours of video into mobile-ready clips with AI

Glance, a mobile-first content platform, uses AI to transform long-form videos into 30-second to 3-minute vertical clips optimized for mobile lock screens. With AI-powered video generation, Glance can process thousands of hours of video content daily, identifying key moments, detecting speakers, and preserving context. The solution uses Google Cloud Speech-to-Text, Gemini, and Google Vision API, combined with custom video manipulation tools. This enables Glance to automate tasks such as caption highlighting, branding, and video reframing. The architecture is divided into three modules: video clipping, speaker detection, and video manipulation.